_ Git - youtube-dl/blob - youtube_dl/extractor/vevo.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5 import xml.etree.ElementTree
   6 import datetime
   7
   8 from .common import InfoExtractor
   9 from ..utils import (
  10     compat_HTTPError,
  11     ExtractorError,
  12 )
  13
  14
  15 class VevoIE(InfoExtractor):
  16     """
  17     Accepts urls from vevo.com or in the format 'vevo:{id}'
  18     (currently used by MTVIE)
  19     """
  20     _VALID_URL = r'''(?x)
  21         (?:https?://www\.vevo\.com/watch/(?:[^/]+/[^/]+/)?|
  22            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
  23            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
  24            vevo:)
  25         (?P<id>[^&?#]+)'''
  26     _TESTS = [{
  27         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  28         'file': 'GB1101300280.mp4',
  29         "md5": "06bea460acb744eab74a9d7dcb4bfd61",
  30         'info_dict': {
  31             "upload_date": "20130624",
  32             "uploader": "Hurts",
  33             "title": "Somebody to Die For",
  34             "duration": 230.12,
  35             "width": 1920,
  36             "height": 1080,
  37         }
  38     }]
  39     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  40
  41     def _formats_from_json(self, video_info):
  42         last_version = {'version': -1}
  43         for version in video_info['videoVersions']:
  44             # These are the HTTP downloads, other types are for different manifests
  45             if version['sourceType'] == 2:
  46                 if version['version'] > last_version['version']:
  47                     last_version = version
  48         if last_version['version'] == -1:
  49             raise ExtractorError('Unable to extract last version of the video')
  50
  51         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
  52         formats = []
  53         # Already sorted from worst to best quality
  54         for rend in renditions.findall('rendition'):
  55             attr = rend.attrib
  56             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
  57             formats.append({
  58                 'url': attr['url'],
  59                 'format_id': attr['name'],
  60                 'format_note': format_note,
  61                 'height': int(attr['frameheight']),
  62                 'width': int(attr['frameWidth']),
  63             })
  64         return formats
  65
  66     def _formats_from_smil(self, smil_xml):
  67         formats = []
  68         smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
  69         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
  70         for el in els:
  71             src = el.attrib['src']
  72             m = re.match(r'''(?xi)
  73                 (?P<ext>[a-z0-9]+):
  74                 (?P<path>
  75                     [/a-z0-9]+     # The directory and main part of the URL
  76                     _(?P<cbr>[0-9]+)k
  77                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
  78                     _(?P<vcodec>[a-z0-9]+)
  79                     _(?P<vbr>[0-9]+)
  80                     _(?P<acodec>[a-z0-9]+)
  81                     _(?P<abr>[0-9]+)
  82                     \.[a-z0-9]+  # File extension
  83                 )''', src)
  84             if not m:
  85                 continue
  86
  87             format_url = self._SMIL_BASE_URL + m.group('path')
  88             formats.append({
  89                 'url': format_url,
  90                 'format_id': 'SMIL_' + m.group('cbr'),
  91                 'vcodec': m.group('vcodec'),
  92                 'acodec': m.group('acodec'),
  93                 'vbr': int(m.group('vbr')),
  94                 'abr': int(m.group('abr')),
  95                 'ext': m.group('ext'),
  96                 'width': int(m.group('width')),
  97                 'height': int(m.group('height')),
  98             })
  99         return formats
 100
 101     def _real_extract(self, url):
 102         mobj = re.match(self._VALID_URL, url)
 103         video_id = mobj.group('id')
 104
 105         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 106         video_info = self._download_json(json_url, video_id)['video']
 107
 108         formats = self._formats_from_json(video_info)
 109         try:
 110             smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 111                 self._SMIL_BASE_URL, video_id, video_id.lower())
 112             smil_xml = self._download_webpage(smil_url, video_id,
 113                                               'Downloading SMIL info')
 114             formats.extend(self._formats_from_smil(smil_xml))
 115         except ExtractorError as ee:
 116             if not isinstance(ee.cause, compat_HTTPError):
 117                 raise
 118             self._downloader.report_warning(
 119                 'Cannot download SMIL information, falling back to JSON ..')
 120
 121         timestamp_ms = int(self._search_regex(
 122             r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
 123         upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
 124         return {
 125             'id': video_id,
 126             'title': video_info['title'],
 127             'formats': formats,
 128             'thumbnail': video_info['imageUrl'],
 129             'upload_date': upload_date.strftime('%Y%m%d'),
 130             'uploader': video_info['mainArtists'][0]['artistName'],
 131             'duration': video_info['duration'],
 132         }